You are on page 1of 22

Development Policy Review, 2007, 25 (2): 243-264

Vulnerability Assessment of Developing


Countries: The Case of Small-island
Developing States
Rosario Turvey
This article puts forward a spatial perspective in framing the methodology
for vulnerability assessment (VA) of developing countries, with special
reference to small-island developing states (SIDS). Geographic
vulnerability from a developing-world perspective is defined by the
countrys susceptibility to physical and human pressures, risks and hazards
in temporal and spatial contexts. In constructing the composite
vulnerability index (CVI), four core indicators are selected as sub-indices.
The study confirms the vulnerability of SIDS based on four dimensions,
namely, coastal index (G1), peripherality index (G2), urbanisation
indicator (G3) and vulnerability to natural disasters (G4), and advocates
consideration of place vulnerability and temporal distinctions when
assessing the vulnerability of SIDS in particular.

Introduction

Over the last few decades, vulnerability assessment has emerged as an alternative
approach to evaluating the situation of developing countries, especially the small-island
developing states (SIDS). The recurring theme of the latters characteristics and distinct
geography is re-examined from the standpoint of their viability, given their
vulnerabilities and uncertainties for survival and sustainability.1
Since the 1990s, there has been a renewed interest in the old problems faced by
SIDS with the emergence of new concerns about vulnerability and sustainable
development. The conventional analysis of small-island characteristics (for example,
remoteness and insularity) as constraint criteria has been supplanted by the emergence
of vulnerability criteria in contemporary research on SIDS. Vulnerability is a multidimensional term that implies a potential for loss from exposure to causal factors such
as biophysical, socio-economic, political and environmental risks and hazards (Cutter,

Assistant Professor in Geography, Department of Geography and Geology, Faculty of Arts and Science,
Nipissing University, 100 College Drive, PO Box 5002, North Bay, ON, Canada P1B 8L7
(rosariot@nipissingu.ca). She wishes to thank the International Development Research Centre (IDRC) for
funding the field research programme in Trinidad and Tobago, Fiji, Samoa and Kiribati and the United
Nations University for the PhD Fellowship with the UN/Institute of Advanced Studies to complete this
research.
1. By underlying geology, structure and altitude, small islands include atolls (living reefs), high islands
(elevated, volcanic, or coral forms), and low-lying islands with a ground surface only a few meters above
the mean sea level (Nunn, 1994). By arbitrary population and area thresholds, small islands have 500,000
or fewer resident population and approximately 10,000 square kilometres or less (Hess, 1990).
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Published by Blackwell Publishing, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.

244

Rosario Turvey

1996; Turner et al., 1996). Broadly, it is implicated in discussions of food security and
hunger, risks and hazards, trade regimes, political economy and global environmental
change (Hewitt and Burton, 1971; Dow, 1992; Downing et al., 1996; Hewitt, 1997;
ECOSOC, 1999).
The purpose of this inquiry stems from the lack of a systematic empirical study
that links geographic theory with vulnerability assessment (VA) of developing
countries, particularly in probing the case of SIDS. The general objective is to explore
the concept of place vulnerability and to develop the VA methodology for measuring
the geographic vulnerability of developing countries. Setting out the VA methodology
involved an empirical research of 100 developing countries, 24 of which are SIDS and
31 Least Developed Countries (LDCs). An approach to VA in geographic terms would
create a better understanding of the geography, environment and development needs of
small-island developing states and other developing countries (UN, 1994).

Review of past studies

In assessing current theory, the literature infers the existence of two theories on smallisland development that deserve careful investigation. One is island development
orthodoxy which refers to the phenomenon of small-island characteristics, such as
remoteness and a narrow resource base as persistent constraints to development. Over
the last two decades, small islands around the world have earned the dubious distinction
of being a special case in view of the tremendous difficulties that beset them in tackling
their environment and development issues (McClean, 1980; Brookfield, 1990; Hess,
1990; UN, 1994). The other theory is small-island vulnerability, as seen in more recent
research. The prevailing conjecture is that small islands are highly vulnerable to various
forces and factors in terms of their precarious geography, fragile economies and
vulnerable environments (UN, 1994; ECOSOC, 1999). Specific factors that give rise to
this vulnerability could be viewed from their vulnerability to environmental problems
due to global warming and sea-level rise and special challenges to planning and
implementing sustainable development.
Despite a burgeoning literature, it is not known whether these two theories are
interrelated in evaluating the situation of developing countries, especially the SIDS;
hence the need for research. Previous empirical research on the vulnerability of SIDS
has been development-based in approach and implemented on an exploratory basis. A
development-based approach to vulnerability aims to reduce the impacts of poverty,
population pressure and the economic forces of globalisation and environmental
degradation. Defined by the UN Committee for Development Policy (1999: 13) as a
risk of being negatively affected by unforeseen events, vulnerability is a relevant
criterion for identifying countries as least developed countries (LDCs).
In this article, developing countries are broadly defined to include SIDS, some of
which are also classified by the UN as LDCs. LDCs are regarded as peripheral
countries and traditional societies in the international political economy, from the
standpoint of their level of development, economic structure and cultural characteristics.
This category of countries was officially instituted in 1971 and is now internationally
recognised as countries accorded special support measures (UNCTAD, 2004). In the
past, LDCs were defined as small, low-income countries that suffer from long-term
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

Vulnerability Assessment of Small-island Developing States 245


handicaps to growth low levels of human resource development and weakness in
economic structure (UN, 1997: 2). Under current UN guidelines, enrolment in the list
of LDCs is not automatic, given a threshold of evaluation criteria of having low-income
(under $750 per capita), weak human assets and economic vulnerability based on
instability of agricultural production and economic smallness, among others (UNCTAD,
2003). Of the current 50 LDCs, 10 are island developing states that fall under the
category of SIDS.
A review of current research on the vulnerability of SIDS found some conceptual
and methodological issues that need to be addressed urgently. First, there is neither a
universally agreed definition nor a clear conception of small-island vulnerability.
Assessing the vulnerability of small-island environments in their geographic
characteristics and island environments could be useful in evaluating their long-term
viability and sustainability. In a survey of current studies, the focus of empirically based
research is on developing a vulnerability index in response to the call by the United
Nations as stipulated in the declaration from the Global Conference on the Sustainable
Development of SIDS (UN, 1994: 51). Secondly, more recent vulnerability studies on
SIDS lack a robust theoretical grounding to inform research in developing a
methodology for vulnerability assessment. This should be addressed if VA is to be used
as an evaluation tool for international development policy and planning on developing
countries.
Since 1995, quantitatively based studies have been carried out to measure the
vulnerability of SIDS. Two thematic areas that dominate the empirically based research
to develop the vulnerability index for small islands are economic and environmental
vulnerability. Economic vulnerability in prevailing scales of analysis is addressed from
the perspective of globalisation and environmental change (Briguglio, 1995; ECOSOC,
1999). Briguglio (1995) provides a perceptive empirical study of the economic
vulnerability of small islands in his analysis of the case of SIDS from the standpoint of
their special disadvantages in developing the vulnerability index. The three indicators
used in the study of 114 countries are export dependence (exposure to foreign economic
conditions), remoteness and insularity and proneness to natural disasters. Though the
results of the study found that SIDS are more vulnerable than any other group of
countries, the index construction has been limited to indicators of economic
vulnerability. As a substantive study, such investigation requires theoretical grounding
in measuring economic vulnerability.
SOPAC (1999) carried out an environmental vulnerability study to construct the
preliminary EVI study for three countries, Australia, Fiji and Tuvalu. There are three
aspects of this type of vulnerability: risks to the environment (natural and
anthropogenic), innate ability of the environment to cope with risks (resilience) and
ecosystem integrity (health or condition of the environment). The index is calculated as
a weighted average of scores in the range of 0 to 7 derived from a total of 57 indicators
that demand a complex process of data generation. Although the study supports the
view that small islands are vulnerable in environmental terms, the results are still
preliminary.
There is potential for re-examining the number and types of variables to be
included in both studies in order to establish clearly which of the causal factors are
considered economic, geographic or ecological determinants of small-island
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

246

Rosario Turvey

vulnerability, as there appears to be some confusion in classifying them. A robust


statement of the theoretical underpinning would increase the effectiveness of research
results according to selected sub-indices. The rigorous process of selecting variables in
constructing the vulnerability index (VI) has been demonstrated but remains
problematic because of insufficient data for certain indicators to be used as sub-indices,
as pointed out in current vulnerability research on developing areas (Briguglio, 1995;
Commonwealth Secretariat, 1997; SOPAC, 1999; ECOSOC, 1999). For this reason, a
more contentious issue in index construction and VA development is the selection of
indicators or sub-indices, because it is contingent on data availability. To date, no
quantitatively based vulnerability assessment on SIDS with a robust theoretical basis
has been undertaken. By broadening the meaning of the term to include geophysical
components, the study asserts that the concepts of place and vulnerability of a place are
bases for understanding small-island vulnerability. Such research should investigate the
links between island development orthodoxy (the special-case argument) and smallisland vulnerability. This should be explored in developing a methodology for the
vulnerability assessment of SIDS in geographic terms. By selecting the sub-indices of
geographic vulnerability that reliably differentiate the most vulnerable countries, the
research could be used as a tool for evaluating the development situation and assistance
needs of developing countries. The findings could be valuable to both the developing
world and the international community in terms of development policy and planning of
countries for inclusion in the list of LDCs.

Vulnerability of place

The methodology for vulnerability assessment is premised upon the concepts of place,
vulnerability of a place and place identity. First, there is a need to situate the concept of
place in vulnerability studies, since it is mostly taken as given. The overarching
objective is to understand why place x is more vulnerable than place y, which is less
vulnerable compared with place z. The term place has basic and mutually constitutive
geographical dimensions like location, size of area and physical environment where
human beings co-exist, based on biophysical, economic and political systems of
interactions. Some authors view it with sensitivity to temporal (time-specific
configurations annual, monthly and seasonal) and geographical variability (local,
national, regional, global) (Watts and Bohle, 1993: 46; Downing et al., 1996). This can
be divided into three geographic elements, namely, (i) social element, meaning those
vulnerable groups or communities living in places at risk or people at risk; (ii) spatial
element, meaning vulnerable places; and (iii) temporal element, specifying time-specific
configurations of the geography of vulnerability.
Secondly, vulnerability of a place is argued as a basic concept in developing the
methodology for VA of developing countries. As listed in Table 1, place vulnerability is
a multiple function of different factors and determinants (economic, geographic and
socio-political) in a given areal or geographical domain (local, state, national and
regional) (Cutter, 1996; Hewitt, 1997). As an evolving concept in geography and
development, it has spatial and temporal aspects and does not operate in a vacuum.

The Author 2007. Journal compilation 2007 Overseas Development Institute.


Development Policy Review 25 (2)

Vulnerability Assessment of Small-island Developing States 247

Table 1: Causal factors of place vulnerability


Economic factors
Vulnerability to external economic shocks
Fragility due to intrinsic factors
Susceptibility of domestic economy to extreme events
Vulnerability to routine risks in everyday life
Geographic factors
Vulnerability to natural disasters
Vulnerability from locational disadvantages peripheries, ghettos, slums
Endangered zones and impaired habitats typhoon and hurricane belts, polluted areas
Vulnerability to structural weakness dependence, food insecurity, powerlessness, poverty
and lack of response capabilities
Fragility of ecosystems and physical environment
Rural, urban, sectoral and communal space of vulnerability
Socio-political factors
Enforced vulnerability of population forced labour, forced resettlement and uprooting,
economic sanctions, ethnic cleansing and regions of misrule
Vulnerability to interpersonal forces population factors, economic, cultural and
environmental conflicts
Sources: Wisner (1993); Watts and Bohle (1993); Briguglio (1995); Hewitt (1997).

Thirdly, there is the need to view the notion of vulnerability in development


geography in the sense of place identity for country classification of developing areas
in the international political economy and the environment. The sense of place identity
in vulnerability is contextual and bound by the ways vulnerability conditions are
produced. SIDS and other developing areas are framed by linking vulnerability into the
classification and structural position of countries in global development. This may mean
a perception of the vulnerability of developing countries as defined by the overall
workings of the political economy in a global system. World countries are often
classified by: (a) level of development (developed and developing, industrialising
economies, least developed); (b) levels of income and extent of indebtedness; and (c)
position in the global economy (North/South axis) (Hewitt, 1997; World Bank, 1999).
Developing countries are invariably categorised by their level of development in
international relations and politics. The question lies not in the number or type of
country classifications, but in terms of why and in what ways distinctions are made.
The distinct category of LDCs is based on the criteria adopted to qualify for special
measures in support of their development needs and to increase their competitive
advantage in the global economy (UNCTAD, 2004). Overall, developing countries are
not homogeneous and have no special treatment, and their country classifications vary
by the nature and purpose of the analysis.
Place identity as it pertains to global environment change is relevant to the case of
SIDS since it is broadly concerned with climate change and global warming issues, and
ecological and technological hazards. This may mean places at risk, the hazards of a
place, or global risk for example, vulnerability to climate change (Cutter, 1996;
Hewitt, 1997). Of greater concern to SIDS (especially the LDCs) is the increasing use
of the so-called graduation approach in the donor process for international
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

248

Rosario Turvey

development co-operation. The idea of graduation means that country x may no longer
be classified as an LDC based on certain criteria established by the UN, and/or a
developing area may no longer be in the donors priority list of developing-country
partners. Graduation from the list of LDCs is perceived to imply change in the countrys
access to markets, aid resources and trading regimes.

The VA methodology

In developing the VA methodology, the study explored geographic vulnerability (GV)


in order to expand the current scope of vulnerability analysis. Past research has
systematically ignored this inherent spatial dimension. GV from a developing-world
perspective is defined by a countrys susceptibility to physical and human pressures,
and risk and hazards in temporal and spatial contexts. Conditions of vulnerability reflect
the complex interaction between the physical and societal systems in a geographic space
and the scale of analysis (i.e., local, regional, global). The inclusion of GV to produce a
Composite Vulnerability Index (CVI) will give a broader view of the issue of
vulnerability by incorporating the geographic environment (i.e., the spatial dimension).
Most crucial to VA development is the investigation of an operationally feasible and
easy-to-use evaluation methodology for those countries facing graduation from LDC
status and those small islands that are economically and environmentally at risk.2 It
should have the capacity to be replicated in similar geographic environments and other
regions of the world for evaluation and country comparisons.
Overall, the criteria adopted in developing the methodology are: (i) simplicity and
ease of application, (ii) capacity for international comparisons, (iii) relevance, (iv)
capability to capture the causal structure of vulnerability, and (v) suitability for
systematic assessment.
The working hypotheses and relationships to be investigated are as follows:
Hypothesis 1: The geographical factors and causal structure are likely to indicate
the vulnerability of a place.
Hypothesis 2: Small islands are likely to be more vulnerable than large developing
countries.
An index of vulnerability is proposed as the measurement of geographic
vulnerability. As such, it is not envisaged as an alternative measure of either growth or
development. Increased concerns about geographical, social, economic and
environmental vulnerability are due to intrinsic factors and external forces, as explored
by past studies. The GV model is an operational framework that is based on place
vulnerability or the sensitivity of an areal or geographic domain in responding to risks,
2. As of the 2003 review, the UN Committee on Development Policy has revised the LDC criteria for
inclusion: Low income, based on a three-year average estimate of the gross national income per capita
(under $750 for inclusion in the LDC list and $900 for cases of graduation); weak human assets based on a
composite Human Assets Index using the indicators of nutrition, health, school enrolment and adult
literacy; and economic vulnerability based on instability of agricultural production, instability of exports
of goods and services, diversification from traditional economic activities and merchandise export
concentration, and economic smallness.
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

Vulnerability Assessment of Small-island Developing States 249


pressures and/or extreme events within a specific timeframe. The GV measurement
areas are: areal factors (geographically localised areas affected by extreme events);
index of exposure to risk; and index of resilience indicating the capacity of the island or
place to cope with risk and recover after the occurrence of natural and anthropogenic
disturbances.

4.1 Index construction


The construction of the index sets out the structure and method of classifying the
sample of developing countries by a GV rank-order system. There are five key steps in
the construction and analysis of the composite vulnerability index (CVI). The process
involves (i) preliminary work; (ii) setting the spatial framework or scope of study; (iii)
construction of the CVI; (iv) conducting vulnerability assessment and analysis; and (v)
reporting and presentation. The first two steps of the preliminary work relate to a review
of the literature, situational analysis and delineation of the study area. The next three
steps specify the actual process of CVI construction, analysis and reporting.
To construct the index, maximum (or minimum) values were taken for each of the
variables. In index calculations and analysis, the question is whether to assign each subindex an equal weight or to give different weights. In practice, equal weighting is the
norm, unless there is a compelling reason for differential weighting of sub-indices. In
this study, the component variables (sub-indices) were equally weighted and a simple
average was then taken to compute the geographic index of vulnerability. Since each of
the variables represents different factors, equal weighting means that the variables were
considered to be equally important. Equal weighting was chosen, in part, based on the
assumption that the four variables represent the slightly different aspects of place
vulnerability. This logic of equal weighting implies that the composite measure (CVI)
represents one dimension of vulnerability (i.e. place vulnerability) as it pertains to
developing areas, with special focus on SIDS.

4.2 Spatial framework


In constructing the CVI, the sample is limited to 100 developing countries (DCs)
representing 59% of total DCs by World Atlas definition. Of this sample, 66 are large
developing countries (or large states) and 34 are small developing countries (or small
states), 24 of which are SIDS. There are 8 SIDS classified as Least Developed
Countries out of the 31 LDCs covered by the study (Tables 2 and 3). Using the
Commonwealth Secretariat (1997) definition, large states have a population of more
than 1.5 million, and small states (including SIDS) 1.5 million or less. The exclusion of
a number of SIDS and other LDCs from the sample was due to insufficiency of data in
all the component indicators.

The Author 2007. Journal compilation 2007 Overseas Development Institute.


Development Policy Review 25 (2)

250

Rosario Turvey

Table 2: SIDS for GV analysis (N=24)


Ocean/developing countries
Atlantic Ocean, n=2
Cape Verde, So Tom and Principe
Caribbean, n= 9
Antigua and Barbuda, Bahamas, Dominica, Grenada, St Kitts and Nevis, St Lucia, St Vincent
and the Grenadines, Jamaica, Trinidad and Tobago
Pacific, n=7
Fiji, Kiribati, Solomon Islands, Tonga, Samoa, Vanuatu, Papua New Guinea
Indian Ocean, n=4
Maldives, Mauritius, Comoros, Seychelles
Mediterranean, n= 2
Cyprus, Malta

Table 3: LDCs for GV analysis (N=31)


LDCs (SIDS and non-SIDS)
SIDS, n=8
Kiribati, Vanuatu, Cape Verde, So Tom and Principe, Solomon Islands, Comoros, Samoa,
Maldives
Non-SIDS, n=23
Mauritania, Zambia, Niger, Bangladesh, Equatorial Guinea, Mozambique, Benin, Haiti,
Uganda, Burkina Faso, Ethiopia, Togo, Sudan, Tanzania, Sierra Leone, Nepal, Rwanda,
Burundi, Madagascar, Democratic Republic of Congo, Lesotho, Myanmar, Gambia

4.3 Organisation of indicators and criteria on CVI


Index and indicators are used in the assessment methodology to capture the underlying
elements of vulnerability.3 In building the set of candidate indicators, a survey was
made of the widely used indicators and a set of criteria was applied to determine the
composite indicators (or sub-indices) of CVI. The organisation of potential indicators
was approached from two directions. First, it placed the indicator selection process
within the concept of place vulnerability, along with other GV concepts as used in the
study to develop the VA methodology. Second, it selected component indicators or subindices from a short list of twelve candidate indicators. All possible relationships
among several indicators and items differing in variance or duplicating one another
were examined to determine the relative strengths of relationships of each candidate
indicator. The criteria applied to select the candidate indicators were validity,
feasibility, plausibility, sensitivity, measurability, clarity and acceptability.

3. ESCAP (1995: 88) defines an indicator as a parameter or a value derived from parameters with a
significance extending beyond that directly associated with it, while an index means a set of aggregated or
weighted parameters or indicators.
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

Vulnerability Assessment of Small-island Developing States 251


The selection of sub-indices is usually done for different purposes. They may be
used to describe and examine a particular situation in a given place, they may be used to
define the relationship between and among component variables, and they may be used
to measure the extent and nature of geographic vulnerability. Other factors are: (i) the
limitation to quantify characteristics that are attributed as causal factors of vulnerability,
and (ii) the availability of data for the sample countries, as reported in standard surveys
and statistical reports on developing countries. In index construction, the final CVI
structure is based on (a) quantity and quality of data, and (b) analytical soundness. The
CVI is a simple average of the four variables G1, G2, G3 and G4, as follows:
CVI =

(G1 + G2 + G3 + G4 )
4

(1)

An indicator that does not contribute to the indexs power is excluded. Data
processed and universally compiled in the field and maintained by a number of data
centres were used as basic data in the index construction, such as UNCTAD trade
statistics and other official reports or a compendium of key indicators. From twelve
candidate indicators, four component variables for the index were chosen to measure
geographic vulnerability. The empirical work in the CVI examined and constructed a
large number of variables to reflect the integral dimensions of geographic vulnerability.
The rank ordering of countries according to a scale of 0-1 indicates those that are most
vulnerable (scale: 0) and those that are least vulnerable (scale: 1). The four sub-indices
or component indicators for CVI are enumerated as follows:
Coastal Index as a proxy for inundation risk (G1). This was calculated by dividing
the length of the coastline (in kilometres) by the total land area of a given country
(Dahl, 1991).4
Peripherality Index (G2): proxy to measure remoteness and insularity, based on
Insurance and Freight Debits as a percentage of Imports of Merchandise in the period
covered: 1990-4, (averaged data).5
Urbanisation Indicator (UR) (G3): URxi/X, expressed as the proportion of
population living in urban areas as expressed in percentages, where URxi is the number
of people living in urban areas and X is the countrys total population.6
Vulnerability to Natural Disasters (G4): Percentage of population affected by
natural disasters during 1970-96; the total number of natural disasters expressed relative
to the total land area against the (natural) logarithm of population (Commonwealth
Secretariat, 1997), period covered: 1970-96.7

4. The empirical work on the CVI examined and constructed a large number of variables to reflect the
integral dimensions of geographic vulnerability.
5. Data from Briguglio (1995) and UNCTAD (1994).
6. Data from Watson et al., 1998; World Bank and World Resources Institute Country at a Glance tables
(1999); United Nations (1995); and Wilkinson and Buddmeier (1994).
7. Database from EM-DAT, Center for Research on the Epidemiology of Disasters (CRED), Catholic
University of Louvain, Brussels; UNCTAD (1995) and Commonwealth Secretariat (1997).
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

252

Rosario Turvey

4.4 Scales and standardisation of vulnerability sub-indices


As done in similar studies, scales and indexes are used as typical ordinal measures and
both involve ranking of variables or units of analysis. As a composite measure, the CVI
is a simple accumulation of scores of each component variable that represents specific
attributes. The scale from 0 to 1 was applied to show the pattern or intensity of
attributes in each of the component indicators.
G1 vulnerability to inundation risk: the most vulnerable to inundation risk in
terms of coastal index (as proxy of inundation risk and elevation factor) receives a
value of 0, the least vulnerable to inundation risk receives a value of 1.
Both the UNEP and the UN documents on vulnerability cited the coastal index as a
measure of insularity. In the context of physical vulnerability, it was used here as a
proxy for the vulnerability of coastal areas to inundation risk, given the inadequate data
such as elevation, i.e., share of land area within 1 metre of mean sea level (UNEP,
1994). As no country-level data sets were available on the mean elevation variable, the
coastal index was used to measure the sensitivity to inundation. Much of the socioeconomic activities in small islands are located along the coastline (Pernetta, 1992).
Data on coastline in kilometres are readily available from the World Resources Institute
and the World Bank database. The GV model used a dummy variable that takes the
value of 1 if the country is landlocked. No island or coastal community is immune to the
potential impacts of climate change, based on projected rates of sea-level rise and global
warming (UNEP, 1994).
More studies are needed on the vulnerability of coastal environments, not only in
relation to climate change issues including sea-level rise, but also to seasonal and interannual climate variability. These represent serious threats to coastal populations and
resources as they tend to experience increased coastal erosion and land loss. In small
islands such as the Bahamas, Kiribati and the Maldives, much of the land area ranges
from only 3 to 5 metres above the present mean sea level (Nurse et al., 1998). Already,
pressures are escalating in coastal areas as rising population, urbanisation in coastal
cities and zones, tourism development and land-based pollution contribute to the
cumulative vulnerability of island environments.
G2 peripherality: the most remote in terms of transport costs (insurance and
freight as percentage of imports) receives a value of 0 and least remote a value of 1.
Remoteness is defined by distance from markets, and physical location is a
permanent characteristic of some small islands and was treated here as a causal factor of
vulnerability. Briguglio (1995) has pointed out that it is remoteness that often creates
economic difficulties and uncertainties. Distance and isolation have produced relatively
high transport costs for a large number of island countries (UN, 1994). Some islands are
constantly bogged down by problems associated with the quality and frequency of
international shipping and air services, and these problems cause undue uncertainties,
delays and higher costs in terms of foreign trade (UN, 1994; Briguglio, 1995). The most
isolated islands face transhipment costs and cessation of transport services (UNCSD,
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

Vulnerability Assessment of Small-island Developing States 253


1998). Most island countries that are LDCs invariably face problems of inadequate
internal transportation, shipping and air transport and other physical infrastructure to
enable them to expand trade activities and increase market access. Devising strategies to
improve transport and communications to remove such development constraints has
posed a major challenge to national governments and the donor community.
G3 vulnerability to urbanisation pressures: the most urbanised in terms of
percentage of urban population receives a value of 0 and the least urbanised a
value of 1.
This indicator is based on the proportion of the population living in urban areas.
Over the past four decades, urbanisation has become one of the most important
geographical phenomena in todays world (Knox et al., 2004: 426). World estimates
indicate that, by 2020, the bulk of the worlds urban population will come from
developing countries. Cities as we know them are magnets to migrant populations in
search of diverse opportunities and places for human settlement. They serve as multiple
settings for human activities, sites for economic interactions and the transformation of
society. By the end of the twenty-first century, more than half of the global population
will live and work in cities or urban environments (Velasquez et al., 1999: 162; UN,
1995). Mega-cities with at least 10 million people have mushroomed in several Asian
developing countries and elsewhere, so it is not surprising to note the twenty-first
century as a century of migration and urbanisation (Knox et al., 2004).
The notion of vulnerability in urban areas suggests the probability of exposure to
pressures and population impacts on urban systems, resources and infrastructure. As a
locus of vulnerability, cities are susceptible to urbanisation pressures from demographic
increases and/or a surge in human activity. According to UNEP (1999), urbanisation is a
pressure indicator of resources and environment, and this is expressed as a proportion
of the population living in urban areas (Earthwatch, 1999). As such, rapid urbanisation
is always associated with tremendous population pressures and locational challenges
ranging from urban congestion to the rise of urban squatter settlements, slums and
public-health risks (Hewitt, 1997; Velasquez et al., 1999). The urban question with
regard to vulnerability also pertains to the capacity of cities to respond to vulnerability
risks and issues due to urban overcrowding, adverse impacts from disasters, threats
from terrorism and other disturbances in everyday urban living.
With regard to SIDS, island populations tend to live in a few urban centres that
include coastal areas where most of the infrastructure and services are located. Past
research reported that damage to critical infrastructure, due to extreme events in urban
settlements with large coastal populations, would be disruptive to some economic,
social and cultural activities (Pernetta, 1992; Nurse et al., 1998). Social and economic
dislocations are more severe in communities with high population densities, particularly
in urban areas (Wilkinson and Buddmeier, 1994). Some island countries have registered
high urban population densities, for example: Euripe, Federated States of Micronesia
(950 per km2); Majuro, Marshall Islands (2,188 per km2). In general, settlements on
atolls and small islands are in highly populated coastal zones and cities in the Pacific
and Indian Oceans, particularly the atoll countries of Kiribati, Tuvalu and Maldives
(Nurse et al., 1998). Since the twentieth century, the rates of urbanised coastal
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

254

Rosario Turvey

populations have increased relative to the demand for economic and environmental
opportunities. By early in the twenty-first century, it is expected that two-thirds of the
population of developing countries will be living along the coast and 70% of the global
human population will live within 60 kilometres of the shoreline (Pernetta and Elder,
1992).
G4 vulnerability to natural disasters: the most vulnerable to natural disasters
receives a value of 0 and the least vulnerable a value of 1.
Since the UN declared 1990-99 as the International Decade for Natural Disaster
Reduction (UN Resolution 42/169, 1987), efforts towards disaster mitigation and
prevention have paralleled efforts to address vulnerability reduction. Poor countries are
hit 500 times more seriously than developed countries, and economic losses are
enormous. Island environments are deemed more vulnerable to tropical storms,
cyclones (hurricanes and typhoons) and drought than the larger masses of land
(McClean, 1980; Brookfield, 1990). Studies by Kay and Hay (1993) indicate that the
southwest Pacific region warmed at a rate of 0.2 C in the 1990s. The perception is that
small islands are vulnerable to natural and environmental disasters because of their
limited capacity to respond and recover from such disasters.
After quantifying each of the component variables, the data sets were then
standardised, since the variables are expressed or measured in different units. Modified
according to Robson (1993) and Briguglio (1995), the formula for standardisation is as
follows:
Vij =

( X ij MaxX i )
( MaxX i MinX i )

i = 1,2,3; j = 1,2,3....100.

(2)

where Vij refers to the degree of vulnerability from the ith variable for country j and Xij
refers to the value of ith variable for country j; MaxXi and MinXi refer to the maximum
and minimum value of the ith variable for all countries in the study. Countries are
generically denoted by j.

Purpose of vulnerability assessment

In essence, the purpose of constructing the GVI and vulnerability assessment is threepronged, namely, for policy evaluation, for spatial analysis in a development context
and for mitigating vulnerability. Each of these purposes is discussed in turn. VA has
potential application in the field and the practice of policy and development evaluation,
in the sense that countries are assessed in terms of their likely vulnerability threats and
risks and their potential to achieve viability in sustaining their national economies.

5.1 Policy evaluation and VA


Broadly, vulnerability studies have become the focus of scientific and international
efforts to address issues about the environment and the sustainability development goals
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

Vulnerability Assessment of Small-island Developing States 255


of developing countries (UN, 1994; 1999). Vulnerability assessment in a policy setting
is a useful tool for country classification and in framing development policies. VA
findings could be used for development evaluation if the task is to set priorities for, and
allocation of, external aid for developing countries. It could also help determine which
country falls in the special or distinct classification or identity, for example, in
identifying countries in the LDC list to be accorded special status in world trade and
frameworks for international co-operation. In the international policy arena, the UN
Committee on Development Policy (1999) has already recognised vulnerability as an
important issue for the developing world, particularly for some SIDS with LDC status.
GVI construction at the national level is meant to operationalise the notion of
vulnerability by taking into account indicators that capture the intrinsic characteristics
whether physical or inherently spatial in nature, namely, aspects of remoteness and/or
environmental susceptibility to natural disasters and hazards. In policy terms, the GVI is
viewed as filling an operational gap in vulnerability studies of interest to the
development policy community as they deal with the question of broadening the
assessment criteria in evaluating the situation of developing countries.
The purpose of constructing a composite vulnerability index is pragmatic, since it
offers an alternative benchmark or reference point for evaluating frameworks for
development policy while building international partnerships and co-operation. In the
practice of evaluation, the conventional approach to measuring development is
commonly based on the economic indicators using Gross Domestic Product (GDP) and
per capita income (PCY) in international dollars by country. Though the economic
criterion is widely used to measure development and classify world countries by income
categories, it does not fully explain a countrys development issues and constraints,
particularly those that persist as intrinsic characteristics of developing countries.
The purpose of VA and index construction has been to explore how GVI and other
statistical measures could be employed as quantitative indicators of the fragility of SIDS
given special reference in this study. If the geographic dimension is considered as an
integral part of vulnerability assessment, it is deemed useful for building the country
vulnerability profile purportedly for policy analysis and planning for development
(UN, 1999). Such country profiles are construed to capture, in a holistic fashion, the
economic, environmental, social and physical dimensions of vulnerability from country
to country. Although the resultant GVI is not itself an index that measures a countrys
growth and development performance, it could broaden our current understanding of the
nature and extent of a countrys vulnerability. What is being posited is that, from a
policy viewpoint, the results from vulnerability assessments could shed light on the
question of vulnerability as it affects the developing world, including SIDS.

5.2 Spatial analysis


VA is straightforward in its application but is deemed a valuable tool for spatial analysis
aimed at stimulating development thinking on the need for global co-operation so that
risks to vulnerability are minimised. When countries are identified and ranked as
vulnerable places, the level of attention is raised concerning their fragile situation or
existing state-of-development in vulnerability terms. In the case of the most vulnerable
countries, there is evidence to suggest that, through the use of VA, the special-case
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

256

Rosario Turvey

argument of a number of SIDS can be seen as appropriate and relevant. In the study, the
results from the paired comparisons of GVI indicate that countries vary in their degree
of vulnerability, due to their geographical position and other such factors that are
structural in nature. Here the structural factors of isolation are spatial and relatively
impervious to policy. The results also suggest that, in establishing their position in the
international development scene, the most vulnerable SIDS, for instance, have to deal
with their vulnerability as a whole, not on a piecemeal basis, in other words, to look at
all dimensions of vulnerability.

5.3 Mitigating vulnerability


Together with the interest in vulnerability index construction as an evaluation tool, the
other equally important task of GVI is to mitigate the risk of vulnerability. Since the
question of vulnerability is not insuperable, the logical step after an assessment is to
determine the level of societal and institutional response system, so that the respective
countries can set out firm decisions and action programmes for reducing the
vulnerability risk. This means careful consideration of the managerial and institutional
capacities of countries in response to vulnerability, and these are often tied to
substantive and procedural processes for adjustment (coping), resistance and recovery
(resilience). This kind of analysis could build on the GVI results through a scenariobuilding exercise that generates the probable conditions of vulnerability and the likely
response pattern or strategies to deal with potential risks, threats and hazards based on
vulnerability assessment.
Whether the assessment results are preliminary or conclusive, a responsemanagement system is essential. Broadly, societal and institutional responses refer to
decisions, strategies and actions taken by groups of people, organisations and agencies
involved in mitigating the risk of vulnerability. An organised response-management
process could include agenda-building and policy and programme development. Other
pertinent actions to reduce the risk of vulnerability are information and knowledgebuilding about vulnerable places, installation and management of community
communication systems for public awareness, early warning signals and alerting,
planning for mitigation, adaptation and emergency preparedness as well as development
of rehabilitation and recovery programmes.

GV results

The results of the CVI construction to measure the index of geographic vulnerability
were categorised into three arbitrary thresholds, constructed to grasp the extent of
vulnerability (from high to low) (see Table 4). Countries in the high GV category are
those that lie between 0 and 0.599 and these refer to countries that are most vulnerable.
Countries in the medium GV category are those between 0.600 and 0.799, while the
countries between 0.800 and 1.000 are the ones that are least vulnerable, based on the
scale of 0-1. Of the 100 developing countries for which the CVI was constructed using
the simple average, 9 are in the high geographic vulnerability (GV) category (0-0.599),

The Author 2007. Journal compilation 2007 Overseas Development Institute.


Development Policy Review 25 (2)

Vulnerability Assessment of Small-island Developing States 257


72 in the medium GV category (0.600-0.799) and 19 in the low GV category (0.8001.000).

Table 4: CVI results by region and GV thresholds


Country
size/region

High GV
(0-0.599)

Medium GV (0.600-0.799)

Low GV
(0.800-1.000)

Large states n=66


Africa

n=1
Democratic
Republic of
Congo

n= 51
Mauritania, Kenya, Algeria, Ethiopia,
Cameroon, Zambia, Niger, Sudan,
Togo, Tanzania, Sierra Leone,
Zimbabwe, Benin, Cte dIvoire,
Nigeria, Rwanda, Uganda, Morocco,
Burkina Faso, Ghana, South Africa,
Egypt, Senegal, Mozambique,
Equatorial Guinea, Congo Democratic
Republic
Paraguay, Bolivia, Haiti, El Salvador,
Venezuela, Peru, Argentina, Dominican
Republic, Uruguay, Brazil, Chile,
Nicaragua, Costa Rica, Mexico,
Ecuador, Colombia, Honduras
Oman, India, Jordan, Bangladesh,
Philippines, Iran, Turkey, Syrian Arab
Republic

n= 14
Tunisia,
Namibia,
Burundi,
Madagascar

South/Central
America

Asia/Pacific/
Others

Guatemala,
Panama

China, Nepal,
Indonesia,
Pakistan,
Thailand,
Myanmar,
Malaysia,
Sri Lanka
n= 5
Cape Verde,
Swaziland,
Lesotho

Small states n=34


Africa

n= 8
So Tom
and Principe

n= 21
Gambia, Botswana, Comoros, Gabon

Caribbean

Bahamas,
Antigua and
Barbuda,
Dominica
Tonga,
Kiribati,
Vanuatu
Seychelles

Trinidad and Tobago, Grenada, St


Vincent , St Kitts and Nevis, St Lucia,
Barbados, Belize, Jamaica, Guyana

Suriname

Fiji, Solomon Islands, Samoa

Papua New
Guinea

Pacific

Indian/
Mediterranean/
Other Asia

Cyprus, Mauritius, Malta, Maldives,


Bahrain

The Author 2007. Journal compilation 2007 Overseas Development Institute.


Development Policy Review 25 (2)

258

Rosario Turvey

Based on the ranking of 100 countries by CVI, the small islands of Tonga,
Bahamas, Kiribati and Vanuatu are the most vulnerable with CVI values of less than
0.599, the threshold for high GV category. Of the 9 countries in the high GV category, 8
are small-island countries, 3 of which are LDCs Kiribati, Vanuatu and So Tom and
Principe. Following the Commonwealth definition, there are 34 small states, of which
24 are SIDS. The island countries found most vulnerable are based on the rank order
system, either with significant peripherality index, urbanisation pressures or
vulnerability to natural disasters.
The result of high GV in one large country (Democratic Republic of Congo) was
also derived in part from values on peripherality and urbanisation pressures. Kiribati
and Seychelles are two SIDS with coastal indices that fall within the values of 0-0.497
in the high GV category. At the other end of the scale, 19 countries are in the low GV
category, 6 of them LDCs with CVI values from 0.812 to 0.920. These countries are
Burundi, Madagascar, Papua New Guinea, Lesotho, Nepal and Myanmar (Table 4).
Apart from Cape Verde, the terrains of the countries in the low GV category are mostly
highland and mountain environments, if not landlocked. If there is any significant
change in one or more of the sub-indices of CVI, those countries in the medium GV
category that include SIDs and LDCs face the risk of sliding past the threshold of the
most vulnerable, i.e. into the high GV category.
By index of peripherality, some LDCs in the low GV category (Burundi,
Madagascar) have sub-index values from 0.264 to 0.564, while other LDCs in the same
category, namely, Myanmar, Nepal, Lesotho, and Cape Verde, have peripherality subindex values from 0.736 to 1.000. There are 22 LDCs in the medium GV category, such
as Maldives (0.615), Mauritania (0.617), Gambia (0.637), Bangladesh (0.0.653),
Solomon Islands (0.690) and Mozambique (0. 695). Mozambique is an example of a
large developing country that has experienced a higher frequency of natural disasters
like India and the Philippines, also in the medium GV category.
Of the most vulnerable SIDS by region, 8 countries are in the Pacific, the
Caribbean and Africa. By country size, 51 large countries have CVI values from 0.600
to 0.799 in the medium GV category, 26 of them developing countries in Africa, 17 in
South/Central America and 8 in Asia/Pacific/Others. Of the 14 large states found to be
least vulnerable, 8 are in Asia/Pacific/Others, 2 are in South and Central America and 4
in Africa. CVI measures the degree or extent of vulnerability and is neither a measure of
economic development nor a measure of poverty. The CVI results confirmed the
hypothesis that small islands and small states including those with high per capita
incomes are more vulnerable than low-income, large countries such as Madagascar. An
issue arising from these results is how to establish the capacity of the most vulnerable
countries by level of resistance and resilience in coping with the degree of vulnerability.
Table 5 shows the vulnerability by country grouping of large states, SIDS and LDCs.
A sensitivity analysis was also conducted to compare the differences between the
results from a simple average and a weighted average of sub-indices. In assigning
weights, G1 and G4 were given a maximum score of 2 each, assuming that these
variables essentially reflect similar aspects of GV, for example susceptibility to natural
disasters, and are considered to be structural factors less affected by the influences and
pressures of policy-making. G2 and G3 are assigned a maximum score of 3 each
because the variables are viewed as more responsive to policy and are therefore treated
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

Vulnerability Assessment of Small-island Developing States 259


as reflecting the different aspects of vulnerability. Based on the weighted averages of
the sub-indices, the results indicated a direct relationship between the sub-indices with
higher weights and the extent of vulnerability. Factors with higher weights (G2 and G3)
contributed most to the increase in a countrys vulnerability and/or CVI ranking.

Table 5: CVI: vulnerability by country grouping


Country grouping

G1

G2

G3

G4

CVI

Large states, n= 66
Small states, n=34
SIDS, n= 24
LDCs, n= 31

0.991
0.862
0.773
0.923

0.600
0.507
0.428
0.421

0.534
0.541
0.612
0.740

0.887
0.783
0.726
0.805

0.758
0.672
0.607
0.721

Notes: G1, Coastal Index, G2, Peripherality Index, G3, Urbanisation, G4, Vulnerability to Natural
Disasters, CVI, Composite Vulnerability Index.

Table 6: Countries of high GV (weighted average)


Developing country
Tonga (SIDS)
Kiribati (SIDS)
Bahamas (SIDS)
Vanuatu (SIDS)
So Tom and Principe (SIDS)
Democratic Republic of Congo
Seychelles (SIDS)
Dominica (SIDS)
Malta (SIDS)
Paraguay
Bahrain
Antigua and Barbuda (SIDS)
Venezuela
Comoros (SIDS)

CVI weighted
average

Country
ranking

CVI simple
average

Country
ranking

0.434
0.460
0.472
0.506
0.511
0.516
0.528
0.540
0.548
0.549
0.558
0.574
0.592
0.596

1
2
3
4
5
6
7
8
9
10
11
12
13
14

0.443
0.496
0.494
0.499
0.557
0.597
0.566
0.579
0.606
0.622
0.622
0.575
0.718
0.653

1
3
2
4
5
9
6
8
10
14
13
7
22
21

Based on the assigned differential weights, countries with high vulnerability to


urbanisation pressures and/or peripherality registered increases and changes in the CVI
scores. Two geographical factors of vulnerability (G2 and G3) are assumed to be more
important and are expected to change within a certain period of time. The number of
countries in the high GV category (0-0.599) increased from 9 to 14 based on the
weighted average of sub-indices, and included 10 SIDS, 2 small states and 2 large
states. By country ranking, the position of 10 countries (for example, Dominica (8),
Equatorial Guinea (29), Kenya (60), Lesotho (98) and Myanmar (100) are unchanged.
Countries such as Burkina Faso, Comoros, and St Lucia showed sharp increases in
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

260

Rosario Turvey

vulnerability due to the effect of one or two sub-indices with assigned weights (Table
6).

Discussion of research findings

The research findings contribute to a growing body of evidence about the vulnerability
of small islands and LDCs. CVI as a measure of geographic vulnerability is a measure
of four sub-indices, namely, coastal index (G1), peripherality index (G2), urbanisation
index (G3) and vulnerability to natural disasters (G4). The CVI results from 100
developing countries indicate how the causal factors expressed as sub-indices produced
geographic vulnerability or defined place vulnerability. The differences in GV
thresholds imply that, although eight of the small islands are in the high GV category,
some large developing countries are also vulnerable as categorised in the high and
medium GV categories. Some large-island countries in the medium GV category are
likewise vulnerable to natural hazards and urbanisation pressures (Table 7).

Table 7: Composite vulnerability index, high GV category


(simple average)
Country
Tonga
Bahamas
Kiribati
Vanuatu
So Tom and Principe
Seychelles
Antigua and Barbuda
Dominica
Democratic Republic of Congo

G1

G2

G3

G4

CVI/rank

0.728
0.881
0.343
0.920
0.903
0.497
0.745
0.908
1.000

0.192
0.724
0.008
0.228
0.100
0.344
0.508
0.516
0.016

0.584
0.046
0.633
0.847
0.562
0.421
0.640
0.253
0.378

0.268
0.324
0.998
0.000
0.682
1.000
0.407
0.639
0.992

0.443 (1)
0.494 (2)
0.496 (3)
0.499 (4)
0.557 (5)
0.566 (6)
0.575 (7)
0.579 (8)
0.597 (9)

The study confirmed the hypothesis that small islands are more vulnerable
compared with large-island countries, based on the four dimensions of geographic
vulnerability as measured by the four sub-indices. In the context of international
development, the inclusion of geographic aspects of vulnerability offers scope for
evaluating the position and situation of SIDS and LDCs. The results from the sensitivity
analysis with differential weights on sub-indices paralleled the CVI scores on simple
averages in terms of the extent of vulnerability. If one country has a high vulnerability
relative to G2 (peripherality) and G3 (urbanisation), vulnerability rises and vice versa.
The results showed that vulnerability increases (or decreases) relative to the sub-indices.
Although the SIDS remained highly vulnerable, the CVI scores showed that large
countries are also highly vulnerable, assuming that peripherality and vulnerability to
urbanisation pressures are more important than vulnerability to inundation risk and
natural disasters. The sensitivity analysis on differential weights found that sub-indices
with higher values have an effect on the overall CVI scores.
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

Vulnerability Assessment of Small-island Developing States 261

Final notes

Overall, the findings from CVI support the need for alternative assessment criteria other
than the limited, conventional economic criteria of per capita income and GDP by
country. With a geographic dimension in vulnerability assessment, building the country
vulnerability profile may prove useful for framing development policy and evaluating
developing countries. While the CVI on GV is not an index of growth and development,
it is useful for understanding the problems of developing countries with respect to
structural constraints and root causes that perpetuate underdevelopment. Eventually, it
could help ascertain why some developing countries need international measures to
remove their persistent constraints on development. More contentious is the possibility
that the most vulnerable LDCs could be among the candidates for graduation from the
LDC list and that they would argue for special treatment, possibly on grounds of
vulnerability. Future work should include the generation of data sets for all sub-indices
to test the validity of baseline and current vulnerability of SIDS. The need for timespecific configurations of geographic variables is to correlate indicators with the extent
or magnitude of risks, pressures and extreme events in order to maintain the viability of
places. To distinguish between baseline and current vulnerability is to determine when a
given situation in a developing area is critical or when a particular place is
geographically vulnerable.8 Baseline vulnerability considers the physical
characteristics and intrinsic properties of a place, the internal and/or external forces, and
the inherent and recurrent factors that may affect, alter or change the condition and
situation of a place. Current vulnerability reflects change in any or all of the component
variables of the baseline vulnerability as a pre-existing parameter. Change may be
indicated by the occurrence of damaging events that create either emergency or nonemergency situations. Furthermore, change may produce critical, short-term conditions
with either temporary or permanent losses, if not irreversible impacts in a place or
developing area.
There are three reasons that support the idea of differentiating baseline from
current vulnerability. First, it measures the temporal and spatial vulnerability
dimensions in order to understand better the causal structure and sources of what and
why a place is geographically vulnerable. Time-specific configurations of geographic
variables are meant to correlate indicators with the extent or magnitude of risks,
pressures and extreme events to sustain the viability of places. Secondly, differentiation
between baseline and current vulnerability is practical, as it separates the temporal
aspects of vulnerability from an emphasis on its chronic, inherent and defining
characteristics (WFP, 1996). The term chronic is used here to mean a recurrent
condition. It is easier to institute response mechanisms if both the baseline and the
current vulnerability are compared. Thirdly, in the context of developing areas, it may
promote regular assessment of their position in order to respond to any significant
change, fluctuation or variation due to pressure(s) from domestic or externally induced
shocks and events. What is urgently needed is a comparative system of vulnerability
assessment regionally and internationally one that differentiates the baseline from the
8. The CVI values are presented as current vulnerability to differentiate the position of each country from the
baseline figures.
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

262

Rosario Turvey

current vulnerability in efforts to programme the assistance needs of developing areas


according to time and space configurations. CVI as a reference point could help assess
the eligibility of developing areas for appropriate international assistance, in other
words, from emergency assistance to programmable foreign aid (for example, the
tsunami disaster in South Asia, in December 2004). Although the GV model on place
vulnerability is straightforward and easy to use, a comparative analysis of baseline and
current vulnerability will be contingent upon the availability of data sets for
international and between-country comparisons.
first submitted September 2005
final revision accepted July 2006

References
Briguglio, L. (1995) Small-island Developing States and their Economic
Vulnerabilities, World Development 23 (9): 1615-32.
Brookfield, H. C. (1990) Introduction to the Conduct and Findings of the InterDisciplinary Fiji Project, in UNESCO, Population-Environment Relations in
Tropical Islands: The Case of Eastern Fiji. MAB Technical Notes 13. Paris:
UNESCO.
Commonwealth Secretariat (1997) A Future for Small States: Overcoming
Vulnerability. London: Commonwealth Secretariat.
Cutter, S. L. (1996) Vulnerability to Environmental Hazards, Progress in Human
Geography 20: 529-39.
Dahl, A. (1991) Island Directory. Nairobi: UNEP Earthwatch, http://unep.ch/
earthwatch/islands (accessed 29 October 1999).
Dow, K. (1992) Exploring Differences in our Common Future(s): The Meaning of
Vulnerability to Global Environment Change, Geoforum 23 (3): 417-36.
Downing, T., Watts, B. and Bohle, H. (1996) Climate Change and Food Security:
Toward A Geography of Vulnerability, in T. Downing (ed.), Climate Change and
World Food Security Vol. 1. NATO ASI Series 37. Heidelberg: Springer-Verlag.
Earthwatch (1999) Island Directory. http://unep.ch/earthwatch, accessed 5 August
1999.
Economic and Social Council of the United Nations (ECOSOC) (1999) Report for
Development Policy. New York: ECOSOC.
Economic and Social Commission for Asia and the Pacific (ESCAP) (1995) State of the
Environment Report: Asia-Pacific Region. Bangkok: ESCAP.
Hess, A. (1990) Overview of Sustainable Development and Environmental
Management of Small Islands, in W. Beller, P. D. dAyala and P. Hein (eds),
Sustainable Development and Environmental Management of Small Islands. Paris:
UNESCO.
Hewitt, K. (1997) Regions of Risk: A Geographical Introduction to Disasters. Harlow:
Longman.
Hewitt, K. and Burton, I. (1971) The Hazardousness of a Place: A Regional Ecology of
Damaging Events. Research Publication No. 6. Toronto: University of Toronto,
Department of Geography.
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

Vulnerability Assessment of Small-island Developing States 263


Kay, R. and Hay, J. (1993) A Decision Support Approach to Coastal Vulnerability and
Resilience Assessment: A Tool for Integrated Coastal Zone Management, in R.
McClean and N. Mimura (eds), Vulnerability Assessment to Sea-Level Rise and
Coastal Zone Management. Proceedings of IPCC Eastern Hemisphere Workshop.
Geneva: IPCC.
Knox, P., Marston, S. and Nash, A. (2004) Human Geography: Places and Regions in
Global Context. 2nd edn. Toronto: Pearson Prentice Hall.
McClean, R. C. (1980) Spatial and Temporal Variability of External Physical Controls
on Small Island Ecosystems, in UNESCO, Population-Environment Relations in
Tropical Islands: The Case of Eastern Fiji. MAB Technical Notes 13. Paris:
UNESCO.
Nunn, P. (1994) Oceanic Islands. Oxford: Blackwell Publishers.
Nurse, L. A., McClean, R. F. and Suarez, A. G. (1998) Small Island States, in
Intergovernmental Panel on Climate Change (IPCC), WMO and UNEP, in The
Regional Impacts of Climate Change: An Assessment of Vulnerability. Cambridge:
Cambridge University Press.
Pernetta, J. C. (1992) Impacts of Climate Change and Rise on Small Island States,
Global Environmental Change 2 (1): 19-31.
Pernetta, J. C. and Elder, D. L. (1992) Climate, Sea and the Coastal Zone: Management
and Planning, Ocean and Coastal Management 18 (1): 113-60.
Robson, C. (1993) Real World Research: A Resource for Social Scientists and
Practitioner-Researchers. Oxford and Cambridge: Blackwell Publishers.
South Pacific Applied Geoscience Commission (SOPAC) (1999) Environmental
Vulnerability Index (EVI). SOPAC Technical Report 275. Report by U. Kaly; L.
Briguglio; H. MacLeod; S. Schmall; C. Platt and R. Pal. Suva: SOPAC.
Turner, R. K., Subak, S. and Adger, W. N. (1996) Pressures, Trends, and Impacts in
Coastal Zones: Interactions Between Socio-economic and Natural Systems,
Environmental Management 20 (2): 159-73.
United Nations (1999) Indicators for Sustainable Development (ISD): Progress from
Theory to Practice. New York: UN Department of Economic and Social Affairs,
www.uncsd.org (accessed 15 October 1999).
United Nations (1997) ESCAP Report: Towards Indicators of Sustainable Development
in Asia and the Pacific. New York: United Nations.
United Nations (1995) Statistics on Population. New York: UN Population Division.
United Nations (1994) Report of the Global Conference on the Sustainable
Development of Small Island Developing States. New York: United Nations.
United Nations Commission for Sustainable Development (UNCSD) (1998) Report of
the Secretary-General on the Development of a Vulnerability Index for SmallIsland Developing States. (Unedited version of the Commission for Sustainable
Development, Sixth Session, 20 April-1 May 1998, and report to the Committee for
Development Policy, 32nd Session, 4-8 May, 1998.) New York: UNCSD.
UN ECOSOC Committee on Development Policy (1999) Report on the First Session,
26-30 April. ECOSOC Supplement No. 13/E. New York: United Nations.
United Nations Conference on Trade and Development (UNCTAD) (2004) Is a Special
Treatment of Small Island Developing States Possible? New York and Geneva:
UNCTAD.
The Author 2007. Journal compilation 2007 Overseas Development Institute.
Development Policy Review 25 (2)

264

Rosario Turvey

UNCTAD (2003) Trade and Environment Review. http://www.unctad.org/en/docs/


ditchted20034a5_en/pdf, accessed 30 August 2005.
UNCTAD (1999) Handbook of International Trade and Development Statistics 19961997. Geneva: UNCTAD.
UNCTAD (1998) The Least Developed Countries 1998 Report. Geneva: UNCTAD.
UNCTAD (1995) Handbook of International Trade and Development Statistics.
Geneva: UNCTAD.
UNCTAD (1994) Handbook of International Trade and Development Statistics 1994.
New York: UNCTAD.
United Nations Environmental Programme (1999) Problems in the Small Island
Environment: Small Island Environmental Management. New York: UNEP,
www.unep.ch/islands (accessed January 2000).
UNEP (1994) Earthwatch: Island Directory. http://unep.ch/earthwatch/islands
(accessed 29 October 1999).
Velasquez, G. T.; Uitto, J. I.; Wisner, B. and Takashashi, T. (1999) A New Approach
to Disaster Mitigation and Planning in Mega-cities: The Pivotal Role of Social
Vulnerability in Disaster Management, in T. Inoguchi, E. Newman and G.
Paoletto (eds), Cities and the Environment: New Approaches for Eco-societies.
Tokyo: United Nations University Press.
Watson, R. T., Zinyowera, M. C. and Moss, R. (1998) The Regional Impacts of Climate
Change, An Assessment of Vulnerability: Report of the Intergovermental Panel on
Climate Change. Cambridge: Cambridge University Press.
Watts, M. and Bohle, H. (1993) The Space of Vulnerability: the Causal Structure of
Hunger and Famine, Progress in Human Geography 17: 43-67.
Wilkinson, C. R. and Buddmeier, R. W. (1994) Global Climate Change and Coral
Reefs: Implications for People and Reefs. Report of the IOC-ASPEI-IUCN Global
Task Team on the Implications of Climate Change on Coral Reefs, International
Union for the Conservation of Nature and Natural Resources (IUCN). Gland:
IUCN.
Wisner, B. (1993) Disaster Vulnerability, Geographical Scale and Existential Reality,
in H. G. Bohle (ed.), Worlds of Pain and Hunger: Geographical Perspectives on
Disaster Vulnerability and Food Security. Fort Lauderdale, FL: Freiburg Studies
in Development Geography.
World Bank (1999) World Development Report 1999/2000. New York: Oxford
University Press for the World Bank, http://www1.worldbank.org (accessed 12
November 2000).
World Food Programme (1996) WFP Vulnerability Mapping Guidelines.
www.wfp.it/vam (accessed 15 October 1999).
World Resources Institute (1995) Environmental Indicators: A Systematic Approach to
Measuring and Reporting on Environmental Policy Performance in the Context of
Sustainable Development. Washington, DC: WRI.

The Author 2007. Journal compilation 2007 Overseas Development Institute.


Development Policy Review 25 (2)

You might also like